スキル agent-fs
📦
agent-fs
コンテンツリビジョン r1 中リスク ⚙️ 外部コマンド🌐 ネットワークアクセス📁 ファイルシステムへのアクセス🔑 環境変数
Manage Versioned Files with agent-fs
Agent workflows need persistent files that remain searchable and traceable across changes. This skill guides agent-fs operations for storage, search, version history, collaboration, and SQL queries.
対応: Claude Codex Code(CC)
自分のエージェントでインストール
このリクエストをエージェントにコピーしてください。正規の Skill ページとマニフェストが含まれています。
エージェントリクエスト
Review the Skillstore skill "agent-fs" from https://skillstore.io/skills/desplega-ai-agent-fs.md and its manifest at https://skillstore.io/api/skills/desplega-ai-agent-fs/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.エージェントは引き続き計画を提示し、セキュリティポリシーで必要な確認を求める必要があります。
エージェントが読めるリソース
AI エージェント、クローラー、スクリプトがページ全体ではなく整理されたコンテキストを必要とする場合は、これらのリンクを使ってください。
テストする
「agent-fs」を使用しています。 Explain how sharing a report differs between S3 and local storage.
期待される結果:
- S3 produces a presigned link that anyone holding it can use until expiration.
- Local storage produces an authenticated application link that requires drive membership and does not expire.
- Use the shortest practical expiration for presigned links and keep them out of logs.
「agent-fs」を使用しています。 How should I update a document without overwriting a teammate's changes?
期待される結果:
- Confirm the organization and drive, then read the document and its current version.
- Submit the update with the expected version and a descriptive version message.
- If the version check fails, review the newer content before retrying.
セキュリティ監査
中リスクv1 • 2026/9/28 バージョン別レポートを開く
Most static findings are Markdown delimiters, documented application operations, or scoped authentication examples rather than malicious behavior. Six findings identify risky FUSE permission changes and forced system configuration replacement; one additional finding covers S3 credentials exposed through command arguments. No evidence found of prompt injection or intentional credential exfiltration; external CLI implementations were not included in this review.
1
スキャンされたファイル
495
解析済み行数
6
レビュー項目
0
誤検知を無視
確認済みのセキュリティ上の懸念 (1)
中
S3 Onboarding Exposes Credentials Through Command Arguments
The onboarding example passes credentials through --s3-access-key and --s3-secret-key without a handling warning. Following it can expose keys in process arguments, shell history, or recorded agent commands.
The example explicitly places both credentials in command arguments. Unlike the remote-mount guidance, this workflow gives no warning about argument-based credential exposure.
機能レビュー項目 (6)
これらは、このスキルに期待される可能性のある実際のローカル機能であるため、レビューが必要ですが、確認済みの悪意ある動作としてはカウントされません。
高
sudo privilege escalation
sudo chmod 666 /dev/fuse
The unconditional chmod 666 grants every local user device read and write access. It can weaken existing restrictions without checking sandbox isolation.
高
sudo privilege escalation
sudo ln -sf /proc/mounts /etc/mtab
The command force-replaces /etc/mtab as root without inspecting its current state. This can disrupt existing mount-table configuration on systems requiring a different setup.
高
sudo privilege escalation
echo user_allow_other | sudo tee -a /etc/fuse.conf
Appending user_allow_other globally enables unprivileged users to request cross-user FUSE access. The shown mount does not need this policy relaxation.
高
Linux /proc filesystem access
sudo ln -sf /proc/mounts /etc/mtab
The risk is force-replacing /etc/mtab as root with a /proc/mounts link, not reading process secrets. No existing-configuration check protects this system change.
高
Non-standard device file access
sudo chmod 666 /dev/fuse
The device operation makes /dev/fuse writable by all local users. Unconditional permission broadening can defeat a deployment's existing device-access restrictions.
高
Symlink creation
sudo ln -sf /proc/mounts /etc/mtab
The root-level ln -sf can replace an existing /etc/mtab entry without checking whether replacement is appropriate. This risks disrupting host mount-table configuration.
リスク要因
⚙️ 外部コマンド (50)
SKILL.md:36 SKILL.md:38 SKILL.md:39 SKILL.md:41 SKILL.md:45-59 SKILL.md:59-61 SKILL.md:61-63 SKILL.md:63 SKILL.md:67 SKILL.md:68-70 SKILL.md:70-82 SKILL.md:82-84 SKILL.md:84 SKILL.md:85-86 SKILL.md:86-90 SKILL.md:90 SKILL.md:94 SKILL.md:96 SKILL.md:97-106 SKILL.md:106-108 SKILL.md:108-111 SKILL.md:111-113 SKILL.md:113-115 SKILL.md:115 SKILL.md:116-119 SKILL.md:119-127 SKILL.md:127 SKILL.md:128 SKILL.md:129 SKILL.md:130 SKILL.md:131 SKILL.md:132 SKILL.md:133 SKILL.md:134 SKILL.md:135 SKILL.md:136 SKILL.md:137 SKILL.md:138 SKILL.md:139 SKILL.md:140 SKILL.md:142 SKILL.md:148 SKILL.md:149 SKILL.md:150 SKILL.md:152 SKILL.md:158 SKILL.md:159 SKILL.md:160 SKILL.md:161 SKILL.md:162
🌐 ネットワークアクセス (5)
📁 ファイルシステムへのアクセス (13)
監査者: codex
このレポートを共有・引用
バージョン付き評価レポート、中立的なバッジ、埋め込みカード、引用を共有できます。Skillstore は証拠を報告しますが、この Skill が安全かどうかは判断しません。
レポートリンクをコピー
https://skillstore.io/skills/desplega-ai-agent-fs/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdownバッジ
[](https://skillstore.io/skills/desplega-ai-agent-fs?utm_source=security_passport_badge)HTMLバッジ
<a href="https://skillstore.io/skills/desplega-ai-agent-fs?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/desplega-ai-agent-fs/security.svg" alt="Skillstore security assessment" loading="lazy"></a>埋め込みカード
<iframe src="https://skillstore.io/embed/skills/desplega-ai-agent-fs.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>学術引用 (APA · BibTeX · CFF)
APA形式の引用
desplega-ai. (2026). agent-fs security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/desplega-ai-agent-fs/audits/1BibTeX形式の引用
@techreport{desplega-ai-desplega-ai-agent-fs-2026,
author = {desplega-ai},
title = {agent-fs security audit report (audit version 1)},
institution = {Skillstore},
year = {2026},
number = {1},
url = {https://skillstore.io/skills/desplega-ai-agent-fs/audits/1},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "agent-fs security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
- name: "desplega-ai"
date-released: "2026-09-28"
url: "https://skillstore.io/skills/desplega-ai-agent-fs/audits/1"
identifiers:
- type: other
value: "skillstore:desplega-ai-agent-fs:audit:1"
description: "Skillstore immutable audit report identifier"
Skillstore スコア
このスコアの理由 証拠の信頼度: 中 55
アーキテクチャ
85
保守性
87
コンテンツ
65
コミュニティ
91
仕様準拠
作成できるもの
Persist Agent Project Files
Store project documents, check the target drive, and track revisions with version messages and concurrency checks.
Analyze Stored Business Data
Join stored datasets and summarize results with DuckDB while checking row limits and truncation.
Coordinate Shared Documents
Review comments, inspect recent activity, and manage approved drive memberships with explicit roles.
これらのプロンプトを試す
List Files in a Drive
Show my current agent-fs organization and drive, then list files under [folder]. Do not change anything.
Store a Versioned Document
Save [document] to [path] in agent-fs organization [org] and drive [drive]. Check the current version and use a concurrency guard when updating.
Find and Compare Related Documents
Search drive [drive] in organization [org] for [topic]. Summarize relevant files and compare the latest two versions of [path] without modifying them.
Join Stored Datasets
Query [dataset A] and [dataset B] in agent-fs drive [drive] within organization [org]. Join on [key], calculate [metric], and report truncation and assumptions.
ベストプラクティス
- Confirm organization and drive scope before writes, searches, membership changes, or sharing.
- Use expected-version checks for updates and include descriptive version messages.
- Protect credentials and sharing links, and review privileged FUSE setup before approving system changes.
回避
- Running sandbox setup commands unchanged on shared hosts or production systems.
- Putting real credentials into command arguments, shared logs, or agent transcripts.
- Assuming every sharing link is public or every storage backend supports identical expiration behavior.
よくある質問
Can I use this skill with Claude, Codex, or Claude Code?
Yes, the listing supports all three tools. Executing operations requires access to an installed and configured agent-fs CLI.
Do I need Docker or S3?
No. The local filesystem backend stores versioned file bytes on disk without Docker or S3.
Does semantic search work without embeddings?
Semantic results require embeddings. Hybrid search identifies keyword-only results when semantic retrieval is unavailable.
Are generated sharing links private?
S3 presigned links grant access to anyone holding them until expiration. Local-backend links require an authenticated drive member.
What data can I query with SQL?
The guide covers CSV, TSV, Parquet, XLSX, JSON, NDJSON, SQLite, and DuckDB. Database files require table bindings; results have row limits.
Can I mount drives on every operating system?
No. FUSE mounting requires Linux, device access, and system capabilities. Review privileged setup commands before running them.
開発者情報
作成者
desplega-aiライセンス
MIT
Skillstore リビジョン
r1
バージョンに関する注意
作者はバージョンを宣言していません。
参照
bad9dafc37d1638cd29cb9bab06d5f8dbcd0f6c2
メンテナンスの新しさ
2026/9/30
利用状況
0 ダウンロード · 0 閲覧
ファイル構成
📄 SKILL.md